7/25/2019 IDENTIFICAO DE ESPCIES DE CARNVOROS BRASILEIROS (MAMMALIA: CARNIVORA) A PARTIR DE AMOSTRAS DE
1/38
IDENTIFICAO DE ESPCIES DE CARNVOROSBRASILEIROS (MAMMALIA: CARNIVORA) A
PARTIR DE AMOSTRAS DE FEZES UTILIZANDO
SEQNCIAS DE DNAE MICROSCOPIA PTICA
DE PLOS
Vanessa Godoy Graeff
7/25/2019 IDENTIFICAO DE ESPCIES DE CARNVOROS BRASILEIROS (MAMMALIA: CARNIVORA) A PARTIR DE AMOSTRAS DE
2/38
II
PONTIFCIA UNIVERSIDADE CATLICA DO RIO GRANDE DO SUL
FACULDADE DE BIOCINCIAS
PROGRAMA DE PS-GRADUAO EM ZOOLOGIA
IDENTIFICAO DE ESPCIES DE CARNVOROS BRASILEIROS
(MAMMALIA: CARNIVORA) A PARTIR DE AMOSTRAS
DE FEZES UTILIZANDO SEQNCIAS DE DNA E
MICROSCOPIA PTICA DE PLOS
Vanessa Godoy Graeff
Orientador: Dr. Eduardo Eizirik
DISSERTAODEMESTRADO
PORTO ALEGRE RS BRASIL
2008
7/25/2019 IDENTIFICAO DE ESPCIES DE CARNVOROS BRASILEIROS (MAMMALIA: CARNIVORA) A PARTIR DE AMOSTRAS DE
3/38
III
In the end, our society will be defined not only bywhat we create, but by what we refuse to destroy.
John C. Sawhill (1936-2000)
President, The Nature Conservancy, 1990-2000
Dedico queles que um dia
sofrero as conseqncias da insensatez humana.
7/25/2019 IDENTIFICAO DE ESPCIES DE CARNVOROS BRASILEIROS (MAMMALIA: CARNIVORA) A PARTIR DE AMOSTRAS DE
4/38
IV
AGRADECIMENTOS
Muitas pessoas contriburam direta ou indiretamente para que este trabalho fosse realizado. Agradeo
imensamente a todas, mas especialmente...
Ao Prof. Eduardo Eizirik, que mesmo conhecendo as minhas limitaes, aceitou me orientar,transmitindo a paixo pela sua rea de pesquisa e permitindo o comeo de um grande aprendizado
mtuo.
Ao Prof. Nelson Fontoura, pelo espao cedido em seu laboratrio para a triagem do material coletado
em campo.
AosProfs. Tabajara K. Moreira e Cludio A. Mondin, pela identificao das sementes encontradas.
Ao Bil. CristianM. Joenck, pela identificao das penas encontradas.
Bil. Patrcia Hadler (Paleontologia, MCN/FZB) pelo auxlio na identificao de equimdeos.
Ao Prof. Roberto Reis, Jlio Czar Gonzlez e Mrcia M. A. Jardim (FZB), pelo acesso s c olees
mastozoolgicas do Museu de Cincias e Tecnologia (PUCRS) e Museu de Cincias Naturais (FZB).
Ao Geg. Donarte N. dos Santos Jnior e ao Prof. Rgis Lahm (LTIG- PUCRS) pelo auxlio no
geoprocessamento dos dados.
Dra. Laura R.P. Utz, pela assistncia com o material para microscopia ptica.
Raquel Mattos de Oliveira, tcnica do Laboratrio de Biologia Tecidual (PUCRS), pelo auxlio na
digitalizao fotogrfica das lminas de plos.
Aos colegas do Genoma, pela acolhida nesses dois anos e por estarem sempre dispostos a ajudar.
Um agradecimento especial a Cladi, corao e crebro do Genoma, por fazer com muita eficincia o
seu trabalho, permitindo que o nosso seja possvel.
Um muito obrigada ao Bil. Paulo Chaves por toda a ajuda, ensinamentos, imensa pacincia e
amizade no momento em que cheguei de pra-quedas ao Genoma.
A todos os meus eventuais companheiros de campo, que tornaram possvel o desbravamento de reas
remotas e fizeram do trabalho de coleta de fezes algo mais divertido e interessante.
Ao Bil. CristianoM. Marin, presena insubstituvel na minha vida, por ter me acompanhado em vrias
etapas deste projeto, mas principalmente por todo o suporte, amor e companheirismo. Por estar
sempre presente comemorando as minhas conquistas e me ajudando a passar por momentos difceis.
minha famlia, pelo apoio, carinho e compreenso. famlia Minuzzo, que sempre quando possvel
transmite seu carinho e incentivo.
Capes pela bolsa de mestrado concedida.
7/25/2019 IDENTIFICAO DE ESPCIES DE CARNVOROS BRASILEIROS (MAMMALIA: CARNIVORA) A PARTIR DE AMOSTRAS DE
5/38
V
SUMRIO
RESUMO ...............................................................................................................................................VI
APRESENTAO.................................................................................................................................VII
PREFACE .............................................................................................................................................VII
ARTIGO: "DNA sequencing versushair microscopy: a comparison of two methods for the identification
of non-invasive samples from Atlantic Forest sympatric carnivores ....................................................VIII
Abstract.............................................................................................................................................2
Introduction ....................................................................................................................................... 3
Materials and Methods ......................................................................................................................5
Results ..............................................................................................................................................9
Discussion....................................................................................................................................... 11Acknowledgments ...........................................................................................................................16
References......................................................................................................................................16
Figures ..............................................................................................................................................23
Tables ...............................................................................................................................................29
7/25/2019 IDENTIFICAO DE ESPCIES DE CARNVOROS BRASILEIROS (MAMMALIA: CARNIVORA) A PARTIR DE AMOSTRAS DE
6/38
VI
RESUMO
A habilidade para detectar e analisar indcios de animais na natureza, parte integral da
pesquisa e manejo da vida silvestre, se torna fundamental quando a espcie em estudo um
carnvoro. Para as espcies desse grupo, geralmente raras e/ou difceis de capturar, a anlise das
fezes um dos melhores mtodos no-invasivos para a identificao, caracterizao e monitoramento
das populaes. Contudo, identificaes de espcies a partir de amostras fecais podem ser subjetivas
quando baseadas em critrios tradicionais. Neste estudo, comparamos a eficincia de dois mtodos de
identificao de espcies de carnvoros: anlise do DNA fecal e microscopia ptica de plos-guarda
encontrados em fezes. Ambos foram aplicados a 102 amostras de fezes coletadas em uma rea de
Mata Atlntica no Rio Grande do Sul. Atravs da anlise das seqncias de mtDNA obtidasa partirde
70 fezes foi possvel identificar 75,7% destas amostras como pertencentes a 3 espcies de felinos,
21,3% a uma espcie de candeo silvestre e 3% ao co domstico. Plos-guarda foram encontrados
em 56% das fezes coletadas. Atravs da anlise microscpica desses plos identificou-se 55,6% das
amostras em nvel de espcie (trs espcies de feldeos) e 44% em nvel de famlia (Canidae ou
Felidae). Foram analisadas comparativamente 44 amostras sobrepostas pelos dois mtodos.
Desacordos na identificao entre os mtodos ocorreram em apenas trs amostras. No total, 77
amostras de fezes foram identificadas em nvel de espcie por pelo menos um dos mtodos,
permitindo uma caracterizao da dietadestes carnvoros. A identificao baseada em microscopia de
plos requer poucos gastos ou tecnologia, sendo simples e rpida para anlises em campo. Contudo,
algumas caractersticas morfolgicas dos plos-guarda podem influenciar o poder de identificao por
microscopia, em alguns casos podendo levar a uma interpretao errnea ou incompleta dos padres
observados. A identificao baseada na anlise do DNA fecal pode ter alto custo e ser tecnicamente
difcil, mas as informaes que podem ser obtidas atravs deste mtodo superam em quantidade e
qualidade outros mtodos. Assim, considerando as vantagens e desvantagens dos mtodos
analisados, as diferenas entre eles permitem que sejam usados de forma a se complementarem
mutuamente em diversos estudos, propiciando maior acurcia na identificao de espcies a partir de
fezes.
7/25/2019 IDENTIFICAO DE ESPCIES DE CARNVOROS BRASILEIROS (MAMMALIA: CARNIVORA) A PARTIR DE AMOSTRAS DE
7/38
VII
APRESENTAO
O presente trabalho, intitulado Identificao de espcies de carnvoros brasileiros (Mammalia:
Carnivora) a partir de amostras de fezes utilizando seqncias de DNA e microscopia ptica de plos
foi desenvolvido como parte dos requisitos necessrios para a obteno do ttulo de Mestre junto ao
Programa de Ps-Graduao em Zoologia da Pontifcia Universidade Catlica do Rio Grande do Sul.
Este trabalho teve como principais objetivos caracterizar e comparar duas tcnicas de
identificao de espcies de carnvoros a partir de amostras de fezes, avaliando sua adequao e
eficincia para estudos ecolgicos, e aplicando as amostras identificadas em um estudo da
comunidade de carnvoros do CPCN Pr-Mata, RS. Com base na relao entre os resultados obtidos
com informaes j disponveis sobre estas tcnicas, procurou-se contribuir para a consolidao de
metodologias confiveis para a identificao de espcies de carnvoros brasileiros, auxiliando no
desenvolvimento de estratgias adequadas para sua conservao.
Esta dissertao apresentada no formato de um artigo cientfico a ser submetido ao peridico
Journal of Zoology.
PREFACE
The present study, entitled DNA sequencing versus hair microscopy: a comparison of two
methods for the identification of non-invasive samples from Atlantic Forest sympatric carnivores has
been developed as part of the requirementsfor the M.Sc. degreeat the Graduate Program in Zoology
of the Pontifcia Universidade Catlica do Rio Grande do Sul (PUCRS).
The main goals of this project were to characterize and compare two techniques for the
identification of carnivore species based on faecal samples, evaluating theirsuitabilityand efficiency for
ecological studies involving this group, and applying the identified samples in an investigation of the
carnivore community occurring at the Pro-Mata Research Center, Brazil . We also aimed to relate the
results with available information about these techniques, contributing to the elaboration of reliable
protocols for the identification of Brazilian carnivores, which are required for the development ofadequate conservation strategies for these taxa.
This thesis is written as a scientificarticle to be submitted to the Journal of Zoology.
7/25/2019 IDENTIFICAO DE ESPCIES DE CARNVOROS BRASILEIROS (MAMMALIA: CARNIVORA) A PARTIR DE AMOSTRAS DE
8/38
VIII
DNA sequencing versushair microscopy: a comparison of two methods for the
identification of non-invasive samples from Atlantic Forestsympatric carnivores
VANESSA G. GRAEFF, PAULO B. CHAVES, CRISTIANO M. MARIN andEDUARDO EIZIRIK
A ser submetido ao peridico Journal of Zoology
7/25/2019 IDENTIFICAO DE ESPCIES DE CARNVOROS BRASILEIROS (MAMMALIA: CARNIVORA) A PARTIR DE AMOSTRAS DE
9/38
1
DNA sequencing versus hair microscopy: a comparison of two1
methods for the identification of non-invasive samples from Atlantic2
Forest sympatric carnivores3
4
VANESSA G. GRAEFF1, PAULO B. CHAVES1, CRISTIANO M. MARIN1, EDUARDO EIZIRIK1,25
6
7
1 Faculdade de Biocincias, PUCRS, Porto Alegre, Brazil.8
2 Instituto Pr-Carnvoros, Brazil.9
10
11
12
Corresponding author:13
Dr. Eduardo Eizirik14
Faculdade de Biocincias -PUCRS15
Av. Ipiranga 6681, CEP 90619-900 Porto Alegre, RS, Brazil16
Telephone number: 55 (51) 3320.3500 Ext. 472717
Fax number: 55 (51) 3320.361218
Email: [email protected]
20
21
22
Page headings title: Identification of non-invasive samples from Atlantic Forest carnivores23
7/25/2019 IDENTIFICAO DE ESPCIES DE CARNVOROS BRASILEIROS (MAMMALIA: CARNIVORA) A PARTIR DE AMOSTRAS DE
10/38
2
ABSTRACT1
Non-invasive sampling methods provide a means for studying species such as mammalian2
carnivores, which are difficult to survey using traditional techniques. The collection of carnivore faeces,3
a common non-invasivemethod, has been widely used in wildlife ecology studies. However, species4
identification based on fecal samples can be inaccurate or biased, potentially compromisingthe quality5
of all downstream data. We compared the effectiveness of two methods for carnivore species6
identification: faecal DNA analysis and hair microscopy of guard hairsfound in scats. We collected 1027
carnivore scats in a protected area in the Brazilian Atlantic Forest, and submitted these samples to8
parallel DNA- and hair-based analyses. The target mtDNA segment was successfully amplified from9
98% of the samples, and reliable sequence-based identification could be obtained for 70 scats. This10
analysis identified all samples at species level, 75.7% of which belongedto threespecies of wild felids,11
21.3% to a wild canid and 3% to a domestic dog. Guard hairs were found in 57 scats, and microscopic12
analysis of these samples identified ca.56% of them at species level (three felids) and 44% at family13
level (Canidae or Felidae). The comparative analysis used a total of 44 samples that were identified by14
both methods. Identification disagreement between them occurred in only three scat samples. Overall,15
77 samples were identified at species level by at least one of the methods, allowing a dietary16
investigation of the included carnivores, and a more detailed comparison of the food habits of17
Leopardus tigrinus and Puma yagouaroundi. Identification by hair microscopy is a method that has18
proved to be simple, quick and inexpensive. However, some morphological features of guard hairs19
might influence the power of hair-based identification, possibly leading to incomplete or erroneous20
interpretation in some cases. On the other hand, routine DNA-based identification offaeces might be21
costly and technically demanding, but our results support the expectation that it performs better (both in22
terms of quality and quantity of information) than the other method. Overall, it is likely that both methods23
can be applied in a complementary fashion in several sampling situations, with their joint assets24
providing higher accuracy for carnivore faecal identification.25
26
Keywords: faeces, scats, mammals, mitochondrial DNA, DNA barcodes, optical microscopy, fur, dietary27
7/25/2019 IDENTIFICAO DE ESPCIES DE CARNVOROS BRASILEIROS (MAMMALIA: CARNIVORA) A PARTIR DE AMOSTRAS DE
11/38
3
INTRODUCTION1
Mammalian carnivores are often difficult to study because manyare elusive, rare, wide ranging2
or frequently occupy inaccessible areas. The inability to collect sufficient amounts of reliable data on3
carnivore distribution, numbers, population structure, and habitat requirements is a severe impediment4
to the development of effective conservation efforts on behalf of these species, and the lack of reliable5
information often leads to controversies(Johnson et al., 2001).6
Given these difficulties in detecting carnivore species, it is critical to develop new tools to7
improvethe collection of field data. In this regard, the ability to detect and analyze animal sign in the8
wild, an integral part of wildlife research and management, becomes essential when the study focuses9
on one or more carnivorespecies. For this group, faecal analysis is one of the best methods to identify,10
characterize and monitor populations (Foran, Crooks & Minta, 1997). Faecal samples can provide vital11
information on a species distribution and abundance, as well as individual spatio-temporal movement,12
diet and diseases (Kohn & Wayne, 1997). However, unambiguous species identification from scats is13
problematic. Classifications based on traditional criteria such as size and morphology can be quite14
subjective(Davison et al., 2002) and can be confounded by a number of factors, including largeintra-15
specific variability and the occurrenceof sympatric species with similar scats(Foran et al., 1997; Farrell,16
Roman & Sunquist, 2000; Prugh & Ritland, 2005). In addition, in some cases the identification is17
performedbased on the habitat where the sample was found and/or on the food items identified, wh ich18
would lead to circularand possibly erroneous inferences.19
Considering the widespread difficulties in identifying carnivore species based on their scats,20
more reliable methods are required, especially those that can be standardized and cross-compared21
among studies and geographic regions. Two different approaches have been receiving increased22
attention in this regard, namely hair microscopy and molecular genetic techniques, bothof which have23
an interesting potential for aiding in carnivore studies relying upon non-invasivesampling (e.g.scats).24
Scientists working in conservation biology and in ethology are particularly interested in non-invasive25
sampling techniques, as they allow the investigation of various problems without having to capture,26
disturb, or even observe the animal (OBrien, 1996; Taberlet, Luikart & Geffen, 2001; Johnson et al.,27
2001).28
The analysis of hair characteristics in the context of field studies (which may be termed29
ecological trichology) has been applied in several studies of carnivore food habits, in which both30
predator and prey may be identified using macroscopic features or characters visualized under light31
7/25/2019 IDENTIFICAO DE ESPCIES DE CARNVOROS BRASILEIROS (MAMMALIA: CARNIVORA) A PARTIR DE AMOSTRAS DE
12/38
4
microscopy (Teerink, 1991; Quadros & Monteiro-Filho, 2006a) or scanning electron microscopy1
(Chernova, 2002, 2003). There is a growing trend for employing light microscopy to characterize the2
cuticular and/or medullar morphology of guard hairs (a.k.a. overhairs), allowing the development of3
reference collections (Teerink, 1991) and keys for the identification of predators and their prey (e.g.4
Gamberg & Atkinson, 1988; Inagaki & Tsukahara, 1993; Cowell et al., 2001; Chernova, 2002;5
Chernova, 2003; Gonzlez-Esteban, Villate & Irizar, 2006). In the case of predator identification, the6
approach is based on the fact that overhairs may be ingested during self-grooming, and can be found7
later in the animals faeces. A challenge in this case is to isolate the predators hairs amidst a much8
larger universe of prey remains. In addition, it is critical to develop comprehensive reference collections9
and identification keys allowing for a broad standardization of this approach, as well as to evaluate the10
reliability, consistency and discriminatory power of diagnostic morphological characters observed in the11
hairs of different species. In the case of Brazilian mammals, an important contribution in the direction of12
a standardized approach for the use of this technique was provided by Quadros (2002) and Quadros &13
Monteiro-Filho (2006a, 2006b), who developed a reference collection and identification key for multiple14
species, proposedan integrated nomenclatural scheme for hair micro-structural features, and assessed15
the performance of the method in the analysis of carnivore scats collected in one study site.16
With the development of the Polymerase Chain Reaction (PCR) (Saiki et al., 1985), it has17
become possible to use hair, feathers or faeces from free ranging animals as a source of DNA for18
various types of analyses (Hss et al., 1992; Wayne, 1996; Kohn & Wayne, 1997;Johnson et al., 2001;19
Taberlet et al., 2001; Frankham, Ballou & Briscoe, 2004; Prugh et al., 2005). Analysis of scat DNA20
(often referred to as molecular scatology) is a rather recent approach that uses epithelial cells sloughed21
off the colon walland deposited in scatsas a reliable source of DNA (Albaughet al., 1992) to determine22
speciesoriginand several other types of information. By isolating DNA from scats and using molecular23
assays that can be compared against reference samples, it is possible to accurately determine species24
identity, especially with the use of direct DNA sequencing (Paxinos et al., 1997; Wasser et al., 1997;25
Hansen & Jacobsen, 1999; Farrell et al., 2000; Palomares et al., 2002; Miotto et al., 2007).26
The reliability of the various methods employedfor species-level identification of carnivore scats27
has seldom been assessed in a rigorous fashion, especially by directly comparing different approaches.28
Few examples are available of such comparisons, and most have focusedon morphological versus29
molecular methods (e.g. Davison et al., 2002; Prugh & Ritland, 2005). While it seems quite clear that a30
simple assessment of faecal size, shape, and odor will oftenbe error-prone, and is unlikely to provide31
7/25/2019 IDENTIFICAO DE ESPCIES DE CARNVOROS BRASILEIROS (MAMMALIA: CARNIVORA) A PARTIR DE AMOSTRAS DE
13/38
5
the basis for a standardized approach across multiple study sites and carnivore assemblages, little is1
known as to the relative performance of scat DNA versusguard hair microscopy. Both methods seem to2
have their own assets. The hair-microscopy approach is currently less expensive, requires simpler3
laboratory conditions, and may have sufficient power to discriminate all carnivore species, or at least all4
those expected to occur in a given geographic region. In contrast, the DNA-based method is likely to5
have more discriminatory power (as more characters can be surveyed by choosing one or more6
genomic fragments exhibiting high variability), and should be fully amenable to digital standardization7
leading to worldwide sequence data bases of all carnivore species. Given the promise of both8
approaches, it is relevant to address the following questions:9
(i) what is the success rate of each one in the species-level identification of carnivore scats?10
(ii) when applied to the same samples, are both methods congruent in carnivore species assignment?11
(iii) in the case of discrepancies, which method is more likely to be accurate?12
In this study we attempted to address these questions by investigating the species-level13
identification of carnivores occurring in a protected area in the Brazilian Atlantic Forest, employing both14
hair microscopy and DNA sequencing to the same standardized set of field-collected faecal samples.15
We compared the performance of both methods, assessed their congruence, and used the identified16
scat samples to investigate the food habits of this lit tle-studied carnivore community.17
18
MATERIALS AND METHODS19
Study area20
We conducted this study atthe Pro-Mata Research Center (henceforth Pro-Mata RC), a private21
biological reserve located in Rio Grande do Sul state, southernmost Brazil (292851.96S,22
501028.01W) (Figure 1). The climate is mesothermic, with an average annual temperature of 14.5C,23
mildwintersandhotsummers (Bertoletti & Teixeira, 1995), corresponding to Kppen-Geigers Cfazone24
(Peel, Finlayson & McMahon, 2007). The study area comprises approximately 2,994ha ranging from25
600 to900m a.s.l., including plateau areas surrounded by steep slopes leading to adjacent valleys.26
The plateau vegetation is composed of grasslands and natural patches of Araucaria forest (mixed27
ombrophilous forest, dominated by the Neotropical pine Araucaria angustifolia), while the slopes are28
covered with dense ombrophilous forest (Bertoletti & Teixeira, 1995). These two forest landscapes are29
important components of the Atlantic Forest biome (Guedes et al., 2005), harboring a diverse array of30
species, including many endemics. The Atlantic Forest original range covered more than 1.5 million31
7/25/2019 IDENTIFICAO DE ESPCIES DE CARNVOROS BRASILEIROS (MAMMALIA: CARNIVORA) A PARTIR DE AMOSTRAS DE
14/38
6
km2, stretching along a wide latitudinal range and representing the second largest rainforest of the1
Neotropics (Galindo-Leal et al., 2005). Today it represents acritical example of the alarmingsituation of2
most tropical forests in the world, with only 7% of the original area left. It has been recognized as one3
of the 25 worldwide biodiversity hotspots, but its current system of protected areas is still insufficient to4
guarantee the conservation of its biological legacy (Tabarelli et al., 2005). In spite of its relevance and5
critical conservation status, many aspects of this biome are still poorly known, including the composition6
and ecological dynamics of the mammalian community in most of the remaining fragments.7
8
Faecal sampling9
We collected scats opportunistically along roads and trails between September 2006 and July10
2007. A portion of approximately 5 cmin length of each scat was collected for DNA analysis, and was11
thenpreserved in silica desiccantorethanol 96% and stored at -20Cprior to extraction. The remaining12
portion of eachscat was stored in a zip lock bagat -20Cuntil it was processed. We recorded the GPS13
location(UTM system) of eachscatsample, and estimated its age category (old or recent) at the time of14
collection, based upon appearance, exposure of the deposition site, and weather conditions.15
16
DNA-based identification of carnivore species17
Genomic DNA was extracted from all faecal samples by scraping the surface of each scat18
followed by the use of the QIAamp Stool DNA Mini Kit (Qiagen Inc.), according to the manufacturers19
instructions. DNA extractions and subsequent handling of the extracted DNA were carried out in a20
separate laboratory area, in a UV-sterilized laminar flow hood, dedicated to the molecular analysis of21
non-invasivesamples.22
A short segment (ca. 170 bp) of the mitochondrial ATP6 gene was amplified by PCR using23
primer ATP6-DF3 (5-AACGAAAATCTATTCGCCTCT-3) in combination with either ATP6-DR1 (5-24
CCAGTATTTGTTTTGATGTTAGTTG-3) or ATP6-DR2 (5-TGGATGGACAGTATTTGTTTTGAT-3).25
Each 20 l PCR reaction contained 1-9 l of DNA, 1x PCR Buffer (Invitrogen), 2.5 mM MgCl2, 100M26
ofdNTPs, 5 M of each primer and 1.0unit of TaqDNA Polymerase (Invitrogen). The PCR began with27
10 cycles (Touchdown) of 94C for 45s, 60-50C for 45s, and 72C for 90s; this was followed by 3028
cycles of 94C for 45s, 50C for 45s, 72C for 90sand a final extension of 72C for 3min. PCR products29
were stained with GelRed (Biotium),visualized on a 1% agarose gel, sequenced using the DYEnamic30
ET Dye Terminator Sequencing Kit (Amersham Biosciences) and analyzed in a MegaBACE 100031
7/25/2019 IDENTIFICAO DE ESPCIES DE CARNVOROS BRASILEIROS (MAMMALIA: CARNIVORA) A PARTIR DE AMOSTRAS DE
15/38
7
automated sequencer (Amersham Biosciences). Sequences were visually checked using CHROMAS1
2.0(www.technelysium.com.au/chromas.html),manually corrected using MEGA 3.1 (Kumar, Tamura &2
Nei, 2004) and checked against a reference sequence data base of Neotropical carnivores developed3
in our laboratory (Chaves et al., in preparation). Species identification was conducted using simple4
phylogenetic analyses (e.g.the Neighbor-Joining or UPGMA clustering methods based on a matrix of5
either the absolute number of differences or a p-distance) performed with MEGA, since there was6
virtually no within-species variation in the mtDNA segment used here. This procedure was therefore7
simply a test of species-level monophyly, using 1,000nonparametric bootstrap replicates as a measure8
of nodal support.9
10
Species identification using hair microscopy11
Upon arrival in the laboratory, the portion of each scat that had been stored in a zip lock bag12
was placed for 5 days in a glass flask containing a solution composed of 91% ethanol (70%13
concentrated), 5% formaldehyde (10%concentrated), 4% acetic acid, and drops of liquid detergent for14
conservation and sterilization (Mantovani, 2001). The scats were subsequently washed in flowing water15
through a fine (
7/25/2019 IDENTIFICAO DE ESPCIES DE CARNVOROS BRASILEIROS (MAMMALIA: CARNIVORA) A PARTIR DE AMOSTRAS DE
16/38
8
carnivores gathered from two Brazilian museums (Museu de Cincias e Tecnologia PUCRS [MCT-1
PUCRS]and Museu de Cincias Naturais - FZB/RS [MCN-FZB]). The number of samples varied per2
institution and totalled 25 specimens belonging to 15 species and five carnivoran families (3 Cerdocyon3
thous, 1 Conepatus chinga, 1 Conepatus semistriatus, 1 Eira barbara, 1 Galictis cuja, 2 Leopardus4
colocolo, 2 Leopardus geoffroyi, 2 Leopardus pardalis, 1 Leopardus tigrinus, 2 Leopardus wiedii, 25
Lycalopex gymnocercus, 1 Nasua nasua, 2 Procyon cancrivorus, 1 Puma concolor and 3 Puma6
yagouaroundi). Five of these species (C. chinga, C. semistriatus, L. gymnocercus, L. colocoloand L.7
geoffroyi) had not been included in the reference collection assembled by Quadros (2002), so that no8
standard trichological pattern was available for them. To allow for the possibility of detection of non-9
native carnivores, we also generated reference cuticle and medullar slides for domestic dogs ( Canis10
familiaris), using multiple hairs collected from individuals roaming around the Pr-Mata RC11
headquarters (i.e.within the study site).12
13
Dietary analysis14
Prey remains found in the faecal samples (such as bones, teeth, claws,hairs, seeds, arthropod15
fragments, and feathers), were analyzed to investigate the food habits of the carnivore community of16
the Pr-Mata RC. The identification was based on macroscopic and microscopic featuresofthese items17
in comparison with the reference collections of MCT-PUCRS and MCN-FZB. Due to the high18
fragmentation of the mammalian prey remains, it was not possible to use cranial characters for the19
species identification of small mammals, and the analysis was thus concentrated on dental features,20
more specifically on the first molar. For other mammalian groups, characters such as the shape and21
size of teeth, claws and bones were included.22
The importance of each type of prey was analyzed based on its absolute frequency (the23
percentage of scats in which the item was found) as well as its relative frequency (the percentage of24
occurrence of the item relativeto the total number of items). We calculated the food niche breadth for25
each of the identified carnivore species using the standardized Levins index (Krebs, 1998), given as:26
BA=(B-1)/(n- 1)where B is Levinss index (B= 1/ pi2), where pi is the relative frequency of eachitem)27
and n is the number of dietary items. Values approaching 1.0 indicate that the resources are used in28
similar frequencies, while a diet concentrated on some of the items will generate BAvalues close to 0.29
We also calculated the dietary niche overlap betweenthetwospecies for which the sample size30
was largest (the felids Leopardus tigrinus and Puma yagouaroundi), using Piankas (1973) index:31
7/25/2019 IDENTIFICAO DE ESPCIES DE CARNVOROS BRASILEIROS (MAMMALIA: CARNIVORA) A PARTIR DE AMOSTRAS DE
17/38
9
Ojk= pij pik/( p2
ij p2ik)
, where pi is the proportion of item i in the diet of the species j and k,1
respectively (Pianka, 1973). This index ranges from 0 (no overlap) to 1 (complete overlap). To evaluate2
if the estimated niche overlap could have arisen by chance, we tested for statistical significance of the3
calculated value by comparing it to a null distribution generated by randomizing the original data matrix4
(1,000 iterations) using the program EcoSim (Gotelli &Entsminger,2007).5
6
RESULTS7
In the course of 11 field trips we traversed a total of 115 km of dirt roads and forest trails,8
averaging 8 hours or5 km per day, with a mean of 9 scat samples found persurvey (Figure 1). All9
samples selected for collection and analysis were categorized as recent if they were very fresh (i.e.10
deposited in the previous 1-2 days) or old if they were drier and bearing initial signs of fungal growth11
(i.e.likely deposited in the previous 3-7 days). All faeces that appeared to be much older than this latter12
standard (i.e. extremely dry, exhibiting severe fungal growth, or seemingly devoid of fresh organic13
matter) were not collected, and neither were those that had been severely rain-washed. Overall, a total14
of 102 faeces were collected, most of which (49%) were found during the Spring season(September-15
November - Figure 2).16
DNA extraction was performed for all 102 collected scats, and for 98% of them (N= 100) the17
target mtDNA segment could be successfully amplified (this included old scats having experienced18
modest rain-wash and/or showing some fungalcover). High quality DNA sequences could be obtained19
from 73 of these samples, in every case allowing clear species-level identification (Figure 3 and Table20
1). Of these 73 samples, two were identified as having been produced by a domestic dog (Canis21
familiaris), 15by a crab-eating fox (Cerdocyon thous) and 56 by wild felids (eightPuma concolor, 1722
Leopardus tigrinusand 31 Puma yagouaroundi).23
Overhairs were found in 57 of the 102 scats(56%) collected in the field. Microscopic analysis24
of thesehairs identified 32 samples at species level (four felids, see below), 24 only at family level25
(either Canidae or Felidae), and one was left as unidentified as it presented unusual morphological26
features (this sample was subsequently found to originate from a domestic dog on the basis of the DNA27
evidence). Microscopicanalyses of medullar and cuticularpatterns of the hairs collected in museum28
collections showed that each of the investigated species has a unique pattern that can provide29
diagnostic characters, with the exception of the foxes Lycalopex gymnocercusand C. thous. These two30
canid species appear to have essentially identical patterns, precluding their discrimination with the hair31
7/25/2019 IDENTIFICAO DE ESPCIES DE CARNVOROS BRASILEIROS (MAMMALIA: CARNIVORA) A PARTIR DE AMOSTRAS DE
18/38
10
micro-structural characters surveyed in this study. This is a relevant limitation, as both canids are1
known to occur at the study site, and their species-level discrimination is an important requirement for2
detailed ecological studies in the area. The domestic dog hairs analyzed here were quite variable, and3
could not be conclusively distinguished from the two wild canids mentioned above. Therefore the4
identification of any canid samplecould only be confidently achieved at the family level using the hair5
microscopy approach (Table 1). As for the felids, there was a single sample identified as Leopardus6
wiedii, which was subsequently found to have a discrepant identification based on the DNA method7
(see below) and was thus excluded from all downstream analyses (e.g. dietary investigation). The8
remaining samples identified at species level using the trichological approach were all assigned to P.9
concolor, L. tigrinusor P. yagouaroundi(Table1). Figures 4 and 5 show the micro-structural patterns of10
medullaand cuticle of the four carnivore species identified in this study (considering the DNA-based11
identification of C. thousfor some of the analyzed samples).12
Forty-four samples could be identified using both methods, which allowed an assessment of13
their congruence, and a comparison of their performance. There were only three cases of clear14
identification discrepancy between the two methods, one of them involving the felids Leopardus tigrinus15
(DNA) vs.L. wiedii(hair), the second consisting of L. tigrinus(DNA) vs. P. concolor(hair), and the third16
being P. yagouaroundi (DNA) vs. Canidae (hair). These three samples showing evidence of17
identification disagreement were eliminated from all subsequent analyses (e.g.dietary investigation).18
For the remaining 41 samples, both methods agreed in the carnivore identification at the family level.19
However, 20 of these samples (eight canids and 12 felids) could only be identified at species level using20
the DNA-based approach. The two domestic dog samples contained guard hairs, which in one case21
were inconclusive and in the second led to family-level identification as a canid.22
Given the observation that the two methods were largely congruent, and that few instances of23
discrepancy were observed, we pooled the samples identified by both methods to perform a dietary24
analysis of the carnivore species surveyed in this study. A total of 77 samples (excluding the domestic25
dog scats) could be identified at species level by at least one of the methods, allowing for a comparison26
of the food habits of four carnivore species (Table2).27
The most common food item found in thescats samples of C. thouswas insects, followed by28
rodents (see Table 2). L. tigrinus exhibited the most restricted diet, based very strongly on small29
mammals (ca. 90% of the total items), particularly rodents. P. concolor exhibited a varied diet30
composedpredominantly ofvertebrates (including mostly mammals but also a bird and an amphibian),31
7/25/2019 IDENTIFICAO DE ESPCIES DE CARNVOROS BRASILEIROS (MAMMALIA: CARNIVORA) A PARTIR DE AMOSTRAS DE
19/38
11
and P. yagouaroundi also presented a rather generalist feeding ecology, with over 50% of the food1
items consisting of mammals, but other components including various fruits and insects.2
The estimated food niche breadth (B) varied among species, with L. tigrinusfound to have the3
lowest index (3.7) and P. concolorthe highest (10.3). The calculated value for C. thouswas also quite4
high (9.3), and an intermediate value was observed for P. yagouaroundi (6.8). The standardized5
breadthindex (BA) also indicated a broad dietary niche for P. concolor(1.04) and C. thous(0.93), with6
lower values estimated for L. tigrinus(0.68) and P. yagouaroundi(0.44).7
Finally, we performed a more detailed analysis focusing on the two species that presentedthe8
narrowest dietary niche (i.e. identified as those with the most specialized food habits), the small cats9
Leopardus tigrinusand Puma yagouaroundi. These were also the two species with the largest sample10
size (combined n=52), allowing for a more reliable comparison than any other pair of surveyed species.11
An overall assessment indicated that their diet is similar (Piankas overlap index= 0.938), with both12
species relying heavily on small mammals (Table 2). However, this overlap was found not to be13
homogeneous throughout the year, and a seasonalpattern of variation emerged upon further scrutiny14
(Figure 6). Theoverlap index was lowest inthe Fall(0.433) and greatestin the Winter (0.988). Overall,15
it could be observed that L. tigrinus seems to be more specialized in small mammals, particularly16
rodents (a pattern which is maintained throughout the year), while P. yagouaroundi seasonally17
complements this staple resource with insects and fruits.18
19
DISCUSSION20
The success rate of mtDNA amplification from our faecal samples was quite high, above the21
levels reported in several previous studies (e.g.Davison et al., 2002; Prugh & Ritland, 2005; Prugh et22
al., 2005; Smith et al., 2006; Miotto et al., 2007). DNA sequences could be obtained from most of these23
amplified fragments, but only those considered tobehigh quality (n=70) were used in the present study.24
This implies that further effort on repeated attempts atamplification and sequencing of these fragments25
would yield even higher success rates at obtaining reliable DNA sequences from these samples. The26
success rate observed in this study did not seem to be affected by the scat age category, weather27
conditions or deposition site, contrary to reports in previous studies identifying such factors as influential28
for amplification success. It is possible that the field-based triage which excluded the oldest and most29
decomposed scats was sufficient to lead to collection of only faecal samples still containing suitable30
DNA for analysis.31
7/25/2019 IDENTIFICAO DE ESPCIES DE CARNVOROS BRASILEIROS (MAMMALIA: CARNIVORA) A PARTIR DE AMOSTRAS DE
20/38
12
The number of scat samples identified withthe hairmicroscopy approach was lower than that1
achievedwith the DNA-based method. So far, few papershave reported the success rate of employing2
microscopic analysis of hairs found in faeces to identify carnivore species, which limits the comparative3
assessment of our resultsrelative to other studies. Wachter, Jauernig & Breitenmoser (2006), studying4
the diet of Acinonyx jubatus, found that 33%of the scats containedhairs ingested by auto-grooming. In5
another study, Garla, Setz & Gobbi (2001) found only 17.8% of the faecal samples containing hairs6
ingested by Panthera oncaduring its grooming. These lower results might explain the preference for7
microscopic analysis from hairs collected in fur traps (Lynch, Brown & Rochford, 2006; Depue & Ben-8
David, 2007) or shelters (Cowell et al., 2001), as opposed to obtaining them from scats, when such9
approaches are adequate to address the ecological problemat hand.10
Therefore, the ability to isolate the predators own hairs from among a much larger amount of11
prey remains (usually composed mostly offur), seems to be a substantiallimitationofthis approach. In12
addition, the grooming behavior maydiffer among carnivore species or even for a single species across13
seasons, possibly leading to fluctuations in the number of predator hairs found in scats. Another14
obstacle to the complete success of carnivore identification using hair microscopy can be the quality of15
the guard hair itself. In some cases, a missing portion of the shaft may bias the species-level16
identification, if the standardized cuticle pattern occurs only in thatspecific region of the hair. Moreover,17
we observed that some hairs exhibited differences along the shaft, such as the dimension and/orshape18
of scales, possibly misleading the identification based on cuticular patterns. In this study, all the19
samples exhibiting such problems were identified only at the family level. On a positive note, there were20
several cases of complete congruence between the hair- and DNA-based identifications (i.e. for the21
three felids identified in the area), indicating that this method does have a good potential for accurate22
identification in the case of species that consistently exhibit diagnostic characters.23
It is relevant to attempt to understand the causes of the observed cases of identification24
discrepancy between the two methods. Although hairs are known to be very resistant to various25
extrinsic disturbances, it cannot be ruled out that some forces might occasionally alter their cuticular26
pattern. One of the three samples analyzed here that showed a discrepancy between the two27
identification methods might have been caused by such a process. The DNA-basedmethod identified28
this sample as belonging to L. tigrinus, but the hair-basedanalysis positively identified the L. wiedii29
cuticle pattern. The cuticle patterns of these species differonly in the shape but not in the size of the30
scales. Leopardus wiedii has a narrow petal pattern, while L. tigrinus has a narrow diamond petal31
7/25/2019 IDENTIFICAO DE ESPCIES DE CARNVOROS BRASILEIROS (MAMMALIA: CARNIVORA) A PARTIR DE AMOSTRAS DE
21/38
13
pattern(Teerink, 1991). Therefore, it is possible that the cuticle sufferedmechanical friction which could1
have made the scale edges prominent, misleading the hair-based identification in this case. Further2
investigation using controlled comparisons would be important to verify if this situation can indeed3
occur, and how relevant it is in the context of field-based identifications.4
The two other cases of identification discrepancy could not be easily attributed to any natural5
cause. The most likely explanations would involve cross-contamination among samples, which could6
have occurred either with the spurious inclusion of extrinsic hairs in a sample while it was being7
processed, or via undetected contamination in the DNA assays (no contamination was observed in the8
negative controls). The rigorous monitoring of the error rate of both approaches should be an important9
component of all studies employing carnivore faecal analysis, but this is difficult to accomplish when10
only a single method is applied. It may thus be interesting to consider employing both methods for at11
least a sub-sample of the analyzed faeces as a means of independently assessing the identification12
error rate.13
A peculiar case of predator identification involved a sample for which there was no discrepancy14
between the two methods. The sample was identified by the DNA-based analysis as having been15
produced by a Cerdocyon thous, and as a canid by the hair microscopy approach. However, the16
amount of C. thous(i.e.canid) hairs present in the faecal sample was much larger than is usually found17
for the predator as ingested by self-grooming; this suggests that C. thouswas more likely to have been18
a prey than the predator in this case, which would not have been distinguished by the DNA assay . In19
addition, the food item analysis revealed the presence of three molar teeth belonging to a young coati20
(Nasua nasua). A plausible albeit unlikely inference here is that the sample belongs to a young N.21
nasua that might have fed on a fox carcass and also swallowed its own teeth. This sample was22
excluded from the dietary analysis due to this uncertain provenance, but the case illustrates the23
complexity of performing a reliable identification of the predator when the prey is also a carnivore.24
Our dietary analysis results are consistent with some aspects previously reported on the25
feeding ecology of the investigated felid species (Oliveira, 1994; Wang, 2002; Moreno, Kays &26
Samudio, 2006). Surprisingly, despite the knowledge that P. yagouaroundiis a generalist that includes27
fruits and insects in its diet, in this study the dietary analysis shows the large importance of these items28
in the diet of this species during the Fall season. We expected to find this kind of pattern in the diet of C.29
thous, a species that is known to be more of a feeding generalist. Although numerous studies30
demonstrate the importance of fruits and rodents in this species diet (Motta-Junior, Lombardi &31
7/25/2019 IDENTIFICAO DE ESPCIES DE CARNVOROS BRASILEIROS (MAMMALIA: CARNIVORA) A PARTIR DE AMOSTRAS DE
22/38
14
Talamoni, 1994; Juarez & Marinho-Filho, 2002; Gatti et al., 2006), we found a greater presence of1
insects, some rodents and an insignificant proportion of fruits in its diet.2
Another unexpected result was the presence of several intact ticks (Arachnida: Ixodida) in the3
samples. Several studies investigating the feeding ecology of carnivores have not detected any4
evidence of ticks in the scats(e.g. Azevedo et al., 2006; Gatti, 2006; Moreno et al., 2006; Vieira & Port,5
2006; Walker et al.2007). According to Rosalino et al. (2007), two possible origins for the collected6
ticks can be hypothesized: they could be the result of self-grooming behavior, which iscommon among7
several mammals (Seyfarth, 1977; Patenaude & Bovet, 1984; Wilkinson, 1986; Schino, 1988; Mooring,8
Blumstein & Stoner, 2004) including carnivores (Eckstein & Hart, 2000) or they could have been9
ingested as prey ecto-parasites. Assuming that ticks ingested with prey might be found more intact in10
scats than those groomed (since carnivores usually ingest partly intact portions of the prey through11
large bites - Bicknevicius & Valkenburgh, 1996), the unaltered appearance of most of the identified ticks12
suggest that they originate in the prey, and not in the predator itself. All the scat samples containing13
ticks also contained rodents, which may be an interesting indication that these mammals might be the14
source of these ecto-parasites, and that the latter areperhaps particularly abundant in the area.15
In this study, the number of identified carnivore species was smaller than that expected given16
the known carnivore community occurring in the study area. We found tracks of Procyon cancrivorus17
and Eira barbara in the area, and Nasua nasuawas detected as a food item (Table 2), but none of18
these species was identified among the carnivores producing the faecal samples analyzed here. In19
addition,the fox Lycalopex gymnocercushas beendirectly observed in the areaseveral years ago, and20
a camera-trapping study has documented the presence of N. nasua, P. cancrivorus, Leopardus pardalis21
and L. wiedii(Cerveira, 2005). Our results may be influenced by the fact that in some cases only large22
carnivoresdefecate along trails or territorial boundaries (Macdonald, 1980). Additional hypotheses are23
that these species avoid trails due to their movement behavior (N. nasua), potential dog attacks or24
habitat preferences (P. cancrivorus),or use of different territorial marking sites (e.g.on trees, such as25
perhaps L. wiedii andas is known to occur inL. geoffroyi [Johnson & Franklin, 1991]). Inthecase of E.26
barbaraand L. pardalis, it is possible that their densities are low in the area, so that scat deposition27
along the trails by these species maybe sporadic. In the case of the fox L. gymnocercus, the absence28
of detected scat samples may be due to presence of the likely competitor C. thous, which is known to29
be more of a habitat generalist (Juarez & Marinho-Filho, 2002; Vieira & Port, 2006), as well as the30
observed reduction of grassland habitat in the area (Pillar, 2003; Behling et al., 2004; Oliveira & Pillar,31
7/25/2019 IDENTIFICAO DE ESPCIES DE CARNVOROS BRASILEIROS (MAMMALIA: CARNIVORA) A PARTIR DE AMOSTRAS DE
23/38
15
2004). As L. gymnocercusseems to be more restricted to open habitats (Garca & Kittlein, 2005; Vieira1
& Port, 2006), it may be currently absent or occur at lower densities in the surveyed areas (see Figure2
1).3
The identification of carnivore scats using hair microscopy was found to be simple, quick and4
inexpensive, in some cases requiring only one hair per sample. However, according to Gonzlez-5
Esteban et al. (2006) and the results reported here, this method requires guard hairs that arewell-6
formed and complete for a successful application. In addition, there are two types of guard hairs, the7
primaryand secondary hairs (Teerink, 1991). The latter type displaysa somewhat homogeneous hair8
cuticle pattern throughout the hair shaft in all mammalian groups(Meyer, Schnapper & Hlmann, 2002)9
which could become a problem when species identification ofhairs found in scats is aimed. Moreover,10
according to Noback (1951), the structure may differ significantly at different developmental stages of11
the same individual hair. Thesemorphological features might influence the identification power of hair12
microscopy because the guard hairs found in the scats could be of any type or age, and the researcher13
might not be able to distinguish them, thus proceeding with the downstream analyses based on an14
incorrect interpretation of the patterns observed.15
Regarding the molecular analysis, routine DNA identification from faeces might be costly and16
technically demanding, as DNA isknown to be considerably degraded in this type of material, making it17
advisable for these studies to use only relativelyfresh samples (which is not the case when employing18
the hair-based method). However, faecal samples represent an easy way of non-invasivesampling,19
potentially making a large number of individuals available for analysis. The mitochondrial DNA is20
extremely useful for these studies, and as the results presented here demonstrate that it can be21
successfully applied in relatively old and lightly rain-washed samples.22
Thepossibility of reliably identifying carnivore species employingnon-invasivemethods opens23
up new research avenues in the area of carnivore field ecology, and more rigorous and efficient24
approaches should be constantly sought and compared. Our results indicate that the DNA-based25
method applied here was superior to the hair microscopy approach, al though the latter did identify26
samples that the former could not, and there was substantial congruence between them in several27
cases. It will therefore be interesting to further investigate the strengths and limitations of both28
approaches in additional ecological settings, spanning a broader range of field conditions and sampling29
different carnivore species, so as to identify cases in which the use of one or both is particularly30
recommended. The combination of the resolving power of the DNA-based approach with the low cost31
7/25/2019 IDENTIFICAO DE ESPCIES DE CARNVOROS BRASILEIROS (MAMMALIA: CARNIVORA) A PARTIR DE AMOSTRAS DE
24/38
16
and simple requirements of the trichological method may lead to a future strategy that integrates both1
tools, allowing for greater reliability and effectiveness of carnivore identifications in the field.2
3
ACKNOWLEDGMENTS4
The authors would like to thank all the institutions and people who provided thereference material used5
in the hair and dietaryanalysis. We are thankful to T. K. Moreira and C. A. Mondin for help on the seed6
identification process; C. M. Joenck for the feather identification; D. N. Santos Jr. and R. Lahm for7
providing assistance in spatial data processing; L. Utz for the light microscopy material assistance, and8
R. M. Oliveira for the help on the microscopical picture capture of the hairs. We are also grateful to9
Instituto do Meio Ambiente/PUCRS, Instituto Pr-Carnvoros and CAPES for having supported this10
project.11
12
REFERENCES13
Azevedo, F. C. C; Lester, V.; Gorsuch, W.; Larivire, S.; Wirsing, A. J. & Murray, D. L. (2006) Dietary14
breadth and overlap among five sympatric prairie carnivores. J.Zool. (Lond.)269,127-135.15
Albaugh, G. P.; Iyengar, V.; Lohani, A.; Malayeri, M.; Bala, S. & Nair, P. P. (1992) Isolation of exfoliated16
colonic epithelial cells, a novel, non-invasive approach to the study of cellular markers. Int. J.17
Cancer52, 347-350.18
Behling, H.; Pillar, V. D.; Orlci, L. & Bauermann, S. G. (2004) Late Quaternary Araucaria forest,19
grassland (Campos), fire and climate dynamics, studied by high-resolution pollen, charcoal and20
multivariate analysis of the Cambar do Sul core in southern Brazil. Palaeogeogr. Palaeocl.21
Palaeoecol.203, 277-297.22
Bertoletti, J. J. & Teixeira, M. B. (1995) Centro de Pesquisa e Conservao da Natureza Pr-Mata:23
termo de referncia. Divul. Mus. Cinc. Tecnol. (UBEA/PUCRS)2,1-47.24
Bicknevicius, A. R. & Valkenburgh, B. V. (1996) Design for Killing: Craniodental Adaptations of25
Predators. In: Carnivore Behavior, Ecology, and Evolution(J. L. Gittleman, ed.). Cornell University26
Press, New York, 644 pp.27
7/25/2019 IDENTIFICAO DE ESPCIES DE CARNVOROS BRASILEIROS (MAMMALIA: CARNIVORA) A PARTIR DE AMOSTRAS DE
25/38
17
Cerveira, J. (2005) Mamferos silvestres de mdio e grande porte no Planalto Meridional: suas relaes1
com a fragmentao da paisagem e a presena do gado. MSc thesis, Universidade Federal do Rio2
Grande do Sul,Porto Alegre, Brazil.3
Chernova, O. F. (2002) Architectonic and diagnostic significance of hair cuticle. Izv. Akad. Nauk. Ser.4
Biol.29,238-247.5
Chernova, O. F. (2003)Architectonic and diagnostic significance of hair cortex and medulla. Izv. Akad.6
Nauk. Ser. Biol.30,53-62.7
Cowell, D.; Thomas, G.; Lilies, G.; Bradshaw, A.; Midgley, L. & Slater, F. (2001) Monitoring the use of8
artificial log pile otter holts using hair analysis from bedding. IUCN Otter Spec.Group Bull.18,6-11.9
Davison,A.; Birks, J. D. S.; Brookes, R. C.; Braithwaite, T. C. & Messenger, J. E. (2002) On the origin10
of faeces: morphological versusmolecular methods for surveying rare carnivores from their scats.11
J. Zool. (Lond.)257,141-143.12
Depue,J. E. & Ben-David, M. (2007) Hair sampling techniques for river otters. J. Wildl. Mgmt. 71,671-13
674.14
Eckstein, R. A. & Hart, B. L. (2000) Grooming and control of fleas in cats. Appl. Anim. Behav. Sci.68,15
141-150.16
Farrell, L. E.; Roman, J. & Sunquist, M. (2000) Dietary separation of sympatric carnivores identified by17
molecular analysis of scats. Mol. Ecol. 9,1583-1590.18
Foran, D. R.; Crooks & Minta, S. C. (1997) Species identification from scat: an unambiguous genetic19
method. Wildl.Soc.Bull.25,835-839.20
Frankham, R.; Ballou, J. D. & Briscoe, D. A. (2004) Introduction to Conservation Genetics. Cambridge21
University Press, Cambridge, 616 pp.22
Galindo-Leal, C.; Jacobsen, T. R.; Langhammer, P. F. & Olivieri, S. (2005) Estado dos H otspots: A23
Dinmica da Perda de Biodiversidade. In: Mata Atlntica: Biodiversidade, Ameaas e Perspectivas24
(C. Galindo-Leal & I. G. Cmara, eds.). Fundao SOS Mata Atlntica, So Paulo, 472 pp.25
Gamberg, M. & Atkinson, J. L. (1988) Prey hair and bone recovery in ermine scats. J. Wildl. Mgmt. 52,26
657-660.27
7/25/2019 IDENTIFICAO DE ESPCIES DE CARNVOROS BRASILEIROS (MAMMALIA: CARNIVORA) A PARTIR DE AMOSTRAS DE
26/38
18
Garca, V. B. & Kittlein, M. J. (2005) Diet, habitat use, and relative abundance of pampas fox1
(Pseudalopex gymnocercus) in northern Patagonia, Argentina. Mamm. Biol.70,218-226.2
Garla, R. C.; Setz, E. Z. F & Gobbi, N. (2001) Jaguar (Panthera onca) food habits in Atlantic Rain3
Forest of Southeastern Brazil. Biotropica33,691-696.4
Gatti, A. ; Bianchi, R.; Rosa, C. R. X. & Mendes, S. L. (2006) Diet of the crab-eating fox, Cerdocyon5
thous(Carnivora,Canidae) in Paulo Cesar Vinha State Park, Esprito Santo State, Brazil. Mammalia6
153-155.7
Gonzlez-Esteban, J.; Villate, I. & Irizar, I. (2006) Differentianting hair samples of the European mink8
(Mustela lutreola), the American mink (Mustela vison) and the European polecat (Mustela putorius)9
using light microscopy. J. Zool. (Lond.)270,458-461.10
Gotelli, N. J. & Entsminger, G. L. (2007) EcoSim: null models software for ecology.Version 7. Acquired11
Intelligence Inc. & Kesey-Bear,Jericho, Vermont.http://garyentsminger.com/ecosim.htm.12
Guedes, M. L. S.; Batista, M. A.; Ramalho, M.; Freitas, H. M. B. & Silva, E. M. (2005) Breve Incurso13
sobre a Biodiversidade da Mata Atlntica. In: Mata Atlntica e Biodiversidade(C. R. Franke, P. L.14
B. Da Rocha, W. Klein & S. L. Gomes, orgs.) Edufba, Salvador, 476 pp.15
Hansen, M. M. & Jacobsen, L. (1999) Identification of mustelid species: otter (Lutra lutra), American16
mink (Mustela vison) and polecat (Mustela putorius), by analysis of DNA from faecal samples. J.17
Zool. (Lond.)247,177-181.18
Hss, M.; Kohn, M.; Pbo, S.; Knauer, F. & Schrder, W. (1992) Excrement analysis by PCR. Nature19
359,199.20
Inagaki, H. & Tsukahara, T. (1993)A method of identifying chimpanzee hairs in lion feces. Primates34,21
109-112.22
Johnson, W. E. & Franklin, W. L. (1991) Feeding and spatial ecology of Felis geoffroyi in southern23
Patagonia. J. Mamm. 72, 815-820.24
Johnson, W. E.; Eizirik, E.; Roelke-Parker, M. & OBrien, E. J. (2001) Applications of Genetic Concepts25
and Molecular Methods to Carnivore Conservation. In: Carnivore Conservation(J. L. Gittleman, S.26
M. Funk, D. Macdonald & R. K. Wayne, eds.). Cambridge University Press, Cambridge, 675 pp.27
http://garyentsminger.com/ecosim.htm7/25/2019 IDENTIFICAO DE ESPCIES DE CARNVOROS BRASILEIROS (MAMMALIA: CARNIVORA) A PARTIR DE AMOSTRAS DE
27/38
19
Juarez, K.M. & Marinho-Filho, J. (2002) Diet, habitat use, and home ranges of sympatric canids in1
central Brazil.J. Mammal.83,925933.2
Kohn, M. H. & Wayne, R. K. (1997) Facts from feces revisited. Trends Ecol. Evol. 12,223-227.3
Krebs, C. J. (1998) Ecological Methodology. Benjamin/Cummings, Menlo Park, 620 pp.4
Kumar S.; Tamura, K. & Nei, M.(2004) MEGA3: integrated software for molecular evolutionary genetics5
analysis and sequence alignment. Brief.Bioinform5, 150-163.6
Lynch,A. B.; Brown, M. J. F. & Rochford, J. M. (2006) Fur snagging as a method of evaluating the7
presenceand abundance of a small carnivore, the pine marten (Martes martes). J. Zool. (Lond.)8
270,330-339.9
Macdonald, D. W. (1980) Patterns of scent marking with urine and faeces amongst carnivore10
communities. Symp. Zool. Soc. Lond.45,107-139.11
Mantovani, J. E. (2001) Telemetria convencional e via satlite na determinao da rea de vida de trs12
espcies de carnvoros da regio nordeste do estado de So Paulo. PhD thesis, Universidade13
Federal de So Carlos, So Paulo, Brazil.14
Meyer,W.; Schnapper, A. & Hlmann, G. (2002) The hair cuticle of mammals and its relationship to15
functionsof the hair coat. J. Zool. (Lond.)256,489-494.16
Miotto, R. A.; Rodrigues, F. P.; Ciocheti, G. & Galetti, P. M., Jr. (2007) Determination of the minimum17
population size of pumas (Puma concolor) through fecal DNA analysis in two protected Cerrado18
areas in the Brazilian Southeast. Biotropica39,647-654.19
Mooring, M. S.; Blumstein, D. T. & Stoner, C. J. (2004) The evolution of parasite-defence grooming in20
ungulates. Biol. J. Linn. Soc.81,17-37.21
Moreno, R. S.; Kays, R. W. & Samudio, R., Jr. (2006) Competitive release in diets of ocelot ( Leopardus22
pardalis) and puma (Puma concolor) after jaguar (Panthera onca) decline. J.Mammal.87,808-816.23
Motta-Junior, J. C.; Lombardi, J. A.& Talamoni, S. A. (1994) Notes on crab-eating fox (Dusicyon thous):24
seed dispersal and food habits in southeastern Brazil.Mammalia58,156-159.25
Noback, C. R. (1951) Morphology and phylogeny of hair. Ann. N. Y. Acad. Sci.53,476-491.26
7/25/2019 IDENTIFICAO DE ESPCIES DE CARNVOROS BRASILEIROS (MAMMALIA: CARNIVORA) A PARTIR DE AMOSTRAS DE
28/38
20
OBrien, S. J. (1996) Conservation Genetics in the Felidae. In:Conservation Genetics: Cases Histories1
from Nature(J. C. Avise & J. L. Hamrick, eds.). Chapman & Hall, New York, 512 pp.2
Oliveira, J. M. & Pillar, V. D. (2004) Vegetation dynamics on mosaics of Campos and Araucaria forest3
between 1974 and 1999 in Southern Brazil. Comm. Ecol.5, 197-202.4
Oliveira, T. G. (1994) Neotropical Cats: Ecology and Conservation. EDUFMA, So Lus, 220 pp.5
Palomares, F.; Godoy, J. A.; Piriz, A.; OBrien, J. O. & Johnson, W. E. (2002) Faecal genetic analysis to6
determine the presence and distribution of elusive carnivores: design and feasibility for the Iberian7
lynx. Mol. Ecol. 11,2171-2182.8
Patenaude, F. & Bovet, J. (1984) Self-grooming and social grooming in the North American beaver,9
Castor canadensis. Can. J. Zool.62,1872-1878.10
Paxinos, E.; Mcintosh, C.; Ralls, K. & Fleischer, R. (1997) A non-invasivemethod for distinguishing11
among canid species: amplification and enzyme restriction of DNA from dung. Mol. Ecol. 6,483-12
486.13
Peel, M. C.; Finlayson, B. L. & McMahon, T. A. (2007). Updated world map of the Kppen-Geiger14
climate classification. Hydrol. Earth Syst. Sci.11,1633-1644.15
Pianka, E. R. (1973) The structure of lizard communities. Annu. Rev. Ecol. Syst.4,53-74.16
Pillar, V. D. (2003) Dinmica da Expanso Florestal em Mosaicos de Floresta e Campos no Sul do17
Brasil. In: Ecossistemas Brasileiros: Manejo e Conservao(V. Claudino-Sales, Org.). Expresso18
Grfica e Editora, Fortaleza, 391 pp.19
Prugh, L. R. & Ritland, C. E. (2005) Molecular testing of observer identification of carnivore feces in the20
field. Wildl. Soc. Bull.33,189-194.21
Prugh, L. R. & Ritland, C. E.; Arthur, S. M. & Krebs, C. J. (2005) Monitoring coyote population dynamics22
by genotyping faeces. Mol. Ecol.14,1585-1596.23
Quadros, J. 2002. Identificao microscpica de plos de mamferos brasileiros e sua aplicao no24
estudo da dieta de carnvoros. Ph.D. dissertation, Universidade Federal do Paran, Curitiba, Brazil.25
Quadros, J. & Monteiro-Filho, E. L. A. (2006a) Reviso conceitual, padres microestruturais e proposta26
nomenclatria para os plos-guarda de mamferos brasileiros. Rev.Bras.Zool.23,279-292.27
7/25/2019 IDENTIFICAO DE ESPCIES DE CARNVOROS BRASILEIROS (MAMMALIA: CARNIVORA) A PARTIR DE AMOSTRAS DE
29/38
21
Quadros, J. & Monteiro-Filho, E. L. A. (2006b) Coleta e preparao de plos de mamferos para1
identificao em microscopia ptica. Rev.Bras. Zool.23,274-278.2
Rosalino, L. M.; Rodrigues, M.; Santos-Silva, M & Santos-Reis, M. (2007) Unusual findings on host-tick3
interactions throughcarnivore scat analysis. Exp. Appl. Acarol.43,293302.4
Saiki, R. K.; Scharf, S.; Faloona, F.; Mullis, K. B.; Horn, G. T.; Erlich, H. A. & Arnheim, N. (1985)5
Enzymatic amplification of beta-globin sequences and restriction site analysis for diagnosis of sickle6
cell anemia. Science230,1350-1354.7
Schino, G. (1988) Allogrooming as a tension-reduction mechanism: a behavioral approach. Am. J.8
Primatol.16,43-50.9
Seyfarth, R. M. (1977) A model of social grooming among adult female monkeys. J. Theor. Biol.65,10
671-698.11
Smith, D. A.; Ralls, K.; Hurt, A.; Adams, B.; Parker, M. & Maldonado, J. E. (2006) Assessing reliability of12
microsatellite genotypes from kit fox faecal samples using genetic and GIS analyses. Mol. Ecol.15,13
387-406.14
Tabarelli, M.; Pinto, L.P.; Silva, J.M.C.; Hirota, M. & Bede, L. (2005) Challenges and opportunities for15
biodiversity conservation in the Brazilian Atlantic forest. Conserv.Biol.19,695-700.16
Taberlet, P.; Luikart, G. & Geffen, E. (2001) New Methods for Obtaining and Analyzing Genetic Data17
from Free-ranging Carnivores. In: Carnivore Conservation (J. L. Gittleman, S. M. Funk, D.18
Macdonald & R. K. Wayne, eds.). Cambridge University Press, Cambridge, 675 pp.19
Teerink, B. J. (1991) Hair of West-European Mammals. Cambridge University Press, Cambridge, 22420
pp.21
Vieira, E. M. & Port, D. (2006) Niche overlap and resource partitioning between two sympatric fox22
species in southern Brazil. J. Zoll. (Lond.)272,57-63.23
Wachter, B.; Jauernig, O. & Breitenmoser, U. (2006) Determination of prey hair in faeces of free-24
ranging Namibian cheetahs with a simple method. Cat News44,8-9.25
Walker, R. S.; Novaro, A. J.; Perovic, P.; Palacios, R.; Donadio, E.; Lucherini, M.; Pia, M. & Lpez, M.26
S. (2007) Diets of three species of Andean carnivores in high-altitude deserts of Argentina. J.27
Mammal.88,519-525.28
7/25/2019 IDENTIFICAO DE ESPCIES DE CARNVOROS BRASILEIROS (MAMMALIA: CARNIVORA) A PARTIR DE AMOSTRAS DE
30/38
22
Wang, E. (2002) Diets of ocelots (Leopardus pardalis), margays (L. wiedii), and oncillas (L. tigrinus) in1
the Atlantic Rainforest in Southeast Brazil. Stud. Neotrop. Fauna Environ.37,207-212.2
Wasser, S. K.; Houston, C. S.; Koehler, G. M.; Cadd, G. G. & Fain, S. R. (1997) Techniques for3
application of faecal DNA methods to field studies of Ursids. Mol.Ecol.6,1091-1097.4
Wayne, R. K. (1996) Conservation Genetics in the Canidae. In:Conservation Genetics: Cases Histories5
from Nature (J. C. Avise & J. L. Hamrick, eds.). Chapman & Hall, New York, 512 pp.6
Wilkinson, G. S. (1986) Social grooming in the common vampire bat, Desmodus rotundus. Anim.7
Behav.34,1880-1889.8
7/25/2019 IDENTIFICAO DE ESPCIES DE CARNVOROS BRASILEIROS (MAMMALIA: CARNIVORA) A PARTIR DE AMOSTRAS DE
31/38
23
Figure1. Map of the study area showing the different vegetation landscapes occurring at the Pro-Mata RC, So Francisco de Paula Municipality, Rio Grande do Sul State, Brazil, with a
depiction of scat collection points.
7/25/2019 IDENTIFICAO DE ESPCIES DE CARNVOROS BRASILEIROS (MAMMALIA: CARNIVORA) A PARTIR DE AMOSTRAS DE
32/38
24
0
2
4
6
8
10
12
Sept Oct Nov Dec-
Feb
Mar Apr May Jun Jul
Numberofscats
C.thous L.tigrinus P.concolor P.yagouaroundi
Figure 2. Total number of scats collected at Pro-Mata RC, So Francisco de Paula Municipality,
Rio Grande do Sul State,each month over one year of field work. Values presented here
excluded December-February due to the limited sample size.
7/25/2019 IDENTIFICAO DE ESPCIES DE CARNVOROS BRASILEIROS (MAMMALIA: CARNIVORA) A PARTIR DE AMOSTRAS DE
33/38
25
A90A105
A62A79A95A92A85bPya31
A17A27
A37A44A66A65A36A102A100
A42A104A12A24A98A82A13A89A93A101
bPya67
A97
A70A40A7A103
Puma yagouarondi
A80A14bPco34
A50A78A23Pco544
A99A25A32
Puma concolor
A63A64A96bLti74
A8A75
A2A43A20A59A54A10A73A74A83A84A94bLti75
A76
Leopardus tigrinus
CfaCfa2
A91A60
Canis familiaris
A31bCth64
A19A67A38A71A68A55A18A30A26A39A21A22A72A53bCth204
Cerdocyon thous99
99
98
98
99
82
99
Figure 3. Phylogenetic tree of the mtDNA
ATP6gene segment inferred with the UPGMA
method. Sequences labeled A2-A105 are
derived from faecal samples collected in the
Pro-Mata RC. Sequences derived from
voucher specimens are also included (bPya-
P. yagouaroundi; bPco- P. concolor; bLti- L.
tigrinus; Cfa- C. familiaris; bCth- C. thous).
Numbers above branches are bootstrap
values based on 1,000 replications. Carnivore
pictures were modified from Biblioteca Virtual
Luis ngel Arango (2007) and Iwokrama
International Centre for Rain Forest
Conservation and Development (1999).
7/25/2019 IDENTIFICAO DE ESPCIES DE CARNVOROS BRASILEIROS (MAMMALIA: CARNIVORA) A PARTIR DE AMOSTRAS DE
34/38
26
(a)
(b)
Figure 4.Medullar pattern of reference carnivore guard hairs (shield region) enclosed in synthetic balsam: (a) Felidae:
reversed cloisonn with fringed margins; (b) Canidae: anisocellic. The hair pattern nomenclature shown in (a) is based
on Teerink (1991), while that shown in (b) is based on Quadros & Monteiro-Filho, (2006a). Scale bars = 70 m.
7/25/2019 IDENTIFICAO DE ESPCIES DE CARNVOROS BRASILEIROS (MAMMALIA: CARNIVORA) A PARTIR DE AMOSTRAS DE
35/38
27
(a)
(b)
(c)
(d)
Figure 5. Cuticular pattern of reference carnivore guard hairs (shaft region), p rinted on nail varnish: (a) medium
diamond petal pattern ofCerdocyon thous; (b) narrow diamond petal pattern ofLeopardus tigrinus; (c) regular wave
shape ofPuma concolor; (d) large diamond petal pattern ofPuma yagouaroundi. Hair pattern nomenclature is based
on Teerink (1991). Scale bars = 70 m.
7/25/2019 IDENTIFICAO DE ESPCIES DE CARNVOROS BRASILEIROS (MAMMALIA: CARNIVORA) A PARTIR DE AMOSTRAS DE
36/38
28
0%
20%
40%60%
80%
100%
Spring Fall Winter Spring Fall Winter
Leopardus tigrinus Puma yagouaroundi
%s
cats
Rodentia Didelphinae N. nasua
Un. small mammal Insects Fruits
Figure 6.Relative frequency of food items found in 52 faecal samples oftwo small Neotropical cats (L.
tigrinus, N=21; and P. yagouaroundi, N=31) in Pro-Mata RC, Rio Grande do Sul state, southern Brazil.
Values presented here are derived from all samples grouped per season (except for the Summer,
which was excluded from the analysis due to the small available sample size). Ticks and unidentified
vegetable matter were excluded from this analysis, as they were not considered to be nutritional items.
7/25/2019 IDENTIFICAO DE ESPCIES DE CARNVOROS BRASILEIROS (MAMMALIA: CARNIVORA) A PARTIR DE AMOSTRAS DE
37/38
29
Identification method
Hair DNAFamily-or Species-levelidentification (n= 54) (n= 70)
Canidae 18.5% 24.3%
Canis familiaris -- 3.0%
Cerdocyon thous -- 21.3%
Felidae 81.5% 75.7%
Leopardus tigrinus 18.5% 20.0%
Puma concolor 16.7% 11.4%
Puma yagouaroundi 20.4% 44.3%
Table1.Success rate of two methods of species identification (light microscopy of guard hairs and mtDNA
sequencing) applied to 102 carnivore faecal samples collected at Pro-Mata RC, southern Brazil. Three
samples whose identification was discrepant between the two methods were excluded from this comparison
(see text for details).
7/25/2019 IDENTIFICAO DE ESPCIES DE CARNVOROS BRASILEIROS (MAMMALIA: CARNIVORA) A PARTIR DE AMOSTRAS DE
38/38
Cerdocyon thous Leopardus tigrinus Puma concolor Puma yagouaroundi
% scats % items % scats % items % scats % items % scats % itemsFood items (n=14) (n=31) (n=21) (n=30) (n=11) (n=17) (n=31) (n=52)
Mammals
RODENTIA
Sigmodontinae
Un.small size sp. a 35.7 16.1 57.1 40.0 36.4 23.5 58.1 34.6
Un.medium size sp. b 7.1 3.2 43.0 30.0 16.1 9.6
Echimyinae
Phyllomys sp. 7.1 3.2
Erethyzontinae
Sphiggurus villosus 9.1 5.9
Un.small size rodenta 21.4 9.7 4.8 3.3 9.1 5.9 6.4 3.8
DIDELPHIMORPHIA
Didelphinae 19.0 13.3 3.2 1.9
PROCYONIDAE
Nasua nasua 18.2 11.8 6.4 3.8
Un.small size mammal c 28.6 13.0 4.8 3.3 9.1 5.9 9.7 5.7
Un. medium size mammal d 9.1 5.9
Birds 9.1 5.9
Amphibians 9.1 5.9
Insects
Coleoptera 42.9 19.3 9.1 5.9 12.9 7.7
Orthoptera 21.4 9.7 6.4 3.8
Hymenoptera 7.1 3.2 9.1 5.9
Un. larvae 14.3 6.4 3.2 1.9
ArachnidsIxodida 7.1 3.2 9.5 6.7 9.1 5.9 6.4 3.8
Fruits
Arecaceae
Syagarus romanzoffiana 3.2 1.9
Ebenaceae
Diospyros sp. 7.1 3.2 6.4 3.8
Myrtaceae
Feijoa sellowiana 6.4 3.8
Rosaceae
Rubus sp. 3.2 1.9
Un. fruit 3.2 1.9
Vegetable matter 21.4 9.7 4.8 3.3 18.2 11.8 16.1 9.6
Table 2. Food items found in 77 faecal samples from carnivore species at Pro-Mata RC, So Francisco de Paula
Municipality, Rio Grande do Sul state, southern Brazil.
* % scats, percentage of total of scats in which each item was found; % items, percentage of occurrence of each item in relation to the total
number of items. Un., unidentified.
a Small size rodent, mean weight 30 g. Sigmodontinae prey in this range likely belongs to the genus Oligoryzomys(25g) or Akodon(30g).
b Medium size rodent, mean weight > 30 g. Sigmodontinae prey in this range likely belongs to the genus Holochilus(100g).
Top Related